Tuesday, May 08, 2007
Advocating Eiffel at the Regional Free Software Conference
The Regional Free Software Conference is one of the largest Free Software events in Latin America. The 6th and most recent one, in October 2006, was organized at Mendoza, Argentina.
Last year I decided to take there a presentation about Eiffel, oriented to developers who never heard about it, saying mostly "Hey, there is a language you don't know called Eiffel, you can use it, it's cool!". My presentation is available (OpenDocument slides/Powerpoint slides, converted; both are in spanish) are available under a creative commons by-nc-sa license. If you know about Eiffel you will probably understand most of the slides, even if you don't speak spanish. The strong side I perhaps noted most was that Eiffel allows you to have high-level development and efficiency at the same time. They're not the points I like most, but I thought those could get the attention from the public in a 45' talk.
The reaction from the public was interesting. Most questions were about availability of toolkits/libraries (RDBMS, GUI) than for the language itself. Even after years of having these kind of conversations, I still get surprised about how little people care about a language for itself, and how language and library are so hard to split apart. I got the attention for a few people, anyway, more advanced developers mainly: some guys from the python crowd (the python community here is quite large) and a couple of C++ guys that were really hooked (one of them a contributor of Boost).
Still, it's hard to prepare a talk advocating Eiffel for the masses. I want to retry again this year, at the 7th conference; but would like to look for better ways to "sell" Eiffel. So, if any of you had good ideas about what kind of stuff would be nice to present at an "Introducing Eiffel" talk, you will earn a place at my "Thanks!" slide. Thanks! (in advance).
Last year I decided to take there a presentation about Eiffel, oriented to developers who never heard about it, saying mostly "Hey, there is a language you don't know called Eiffel, you can use it, it's cool!". My presentation is available (OpenDocument slides/Powerpoint slides, converted; both are in spanish) are available under a creative commons by-nc-sa license. If you know about Eiffel you will probably understand most of the slides, even if you don't speak spanish. The strong side I perhaps noted most was that Eiffel allows you to have high-level development and efficiency at the same time. They're not the points I like most, but I thought those could get the attention from the public in a 45' talk.
The reaction from the public was interesting. Most questions were about availability of toolkits/libraries (RDBMS, GUI) than for the language itself. Even after years of having these kind of conversations, I still get surprised about how little people care about a language for itself, and how language and library are so hard to split apart. I got the attention for a few people, anyway, more advanced developers mainly: some guys from the python crowd (the python community here is quite large) and a couple of C++ guys that were really hooked (one of them a contributor of Boost).
Still, it's hard to prepare a talk advocating Eiffel for the masses. I want to retry again this year, at the 7th conference; but would like to look for better ways to "sell" Eiffel. So, if any of you had good ideas about what kind of stuff would be nice to present at an "Introducing Eiffel" talk, you will earn a place at my "Thanks!" slide. Thanks! (in advance).
Labels: advocacy, conference, eiffel, talk
posted by Daniel F Moisset @ 02:35 
Comments:
<< Home
It's comic; I am surprised by the remark:"I still get surprised about how little people care about a language for itself, and how language and library are so hard to split apart".
Isn't this logical? How many people would choose for a language, however good it is, when they have to write everything they need, themselves? I do not say this is 100% the case in Eiffel, but I cannot remember one time when I had to do something in C/C++ and I did not have an abundance of libraries available in C to help me doing it. A lot of them even freely available.
Maybe I am too focused on the situation with SmartEiffel, but suppose you have to write an application with a nice GUI and making a connection to, let's assume, PostgreSQL. What are your opions in this case?
OK; Eiffel is great, but time is still money.
Isn't this logical? How many people would choose for a language, however good it is, when they have to write everything they need, themselves? I do not say this is 100% the case in Eiffel, but I cannot remember one time when I had to do something in C/C++ and I did not have an abundance of libraries available in C to help me doing it. A lot of them even freely available.
Maybe I am too focused on the situation with SmartEiffel, but suppose you have to write an application with a nice GUI and making a connection to, let's assume, PostgreSQL. What are your opions in this case?
OK; Eiffel is great, but time is still money.
# posted by 
: 08 May, 2007 14:21
: 08 May, 2007 14:21
Perhaps my comment was misunderstood... I know libraries are really important at the moment of developing real stuff with a language, and I know that it is one of the worst points about Eiffel.
What I was trying to say, it's that evaluating availability of libraries and "goodness" of a language are separate stuff; if the language is bad, software made on it will have vices, no matter how many good libraries you have. If the language is good, you have the chance at least of having just enough libraries to solve your problem, or even deciding to invest on good library development (Time is money, you chose to spend it mantaining a system written with a suboptimal tool, or reimplementing the things you need on a better tool that is lacking it; i am not implying that any choice is obvious because it's not).
And wandering into personal opinions, libraries without contracts are a lot more poor than what they could be, and promote buggy software development.
Post a Comment
What I was trying to say, it's that evaluating availability of libraries and "goodness" of a language are separate stuff; if the language is bad, software made on it will have vices, no matter how many good libraries you have. If the language is good, you have the chance at least of having just enough libraries to solve your problem, or even deciding to invest on good library development (Time is money, you chose to spend it mantaining a system written with a suboptimal tool, or reimplementing the things you need on a better tool that is lacking it; i am not implying that any choice is obvious because it's not).
And wandering into personal opinions, libraries without contracts are a lot more poor than what they could be, and promote buggy software development.
<< Home
